INDEX

Explanations

mathematical questions with integer answers

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 fours

-0.09

amic

-0.07

 Krank

-0.07

sense

-0.06

 whites

-0.06

Ver

-0.06

 sped

-0.06

 Ritch

-0.06

"↵↵

-0.06

POSITIVE LOGITS

-touch

0.11

roulette

0.11

-END

0.11

zaj

0.11

Č↵

0.10

lug

0.10

 Datum

0.10

vÃ©

0.10

rÃ©

0.10

/INFO

0.10

Activations Density 0.001%